This study will continue an examination of the etiology of drug use and drug escalation for adolescents, with special focus on the behavior of black and Hispanic youth. It focuses on the structure, function and composition of adolescent social networks and the changes that occur in those social networks to determine their effect on drug use. Neighborhood and cultural differences are seen as important influences on the type of social networks that are formed and maintained especially in areas of different racial and ethnic composition, and as interacting with social network characteristics in their effect on patterns of drug use. The relationship between social network characteristics and drug use is seen as reciprocal. That is, the characteristics of adolescent social networks are seen as affecting the patterns of drug use and drug use is, in turn, seen as having an effect on the characteristics of a social network. In cooperation with a project funded by the Office of Juvenile Justice and Delinquency Prevention, a sample of 996 adolescents, currently in the eighth and ninth grades in the Rochester City school system was drawn. By oversampling high delinquency areas, a sample that was not only disproportinately represents minority groups but also overrepresents problem drug users has been acquired. This sample is comprised of 69% black youth, 17% Hispanic, and 14% white. Students and one of their parents or care-taking guardians are interviewed at six month intervals. The first wave of student and parent interviews that contain social network and an extended battery of drug use items is on schedule and nearing completion with 90% of the student interviews and 60% of the parent interviews completed. Additional information from teacher questionnaires, school records, and criminal justice records will be collected within the next few months. Funding provided by the current NIDA grant will allow for more waves of data collection. Respondents will be in their ninth and tenth grade years by the third wave of interviews. Based on preliminary data analysis, drug use among this group is still at the beginning stages. The current proposal would extend data collection through six waves as originally planned, when the adolescents will be at the end of their 10th and 11th grades. This will allow for an examination of the maintenance and escalation of drug use, and to study the effect of changes in adolescent social networks as they affect and are affected by drug use.